DEVELOPMENT OF AN INTELLIGENT ROBOT-ASSISTED SYSTEM FOR FEMUR FRACTURE ALIGNMENT USING A STEWART PLATFORM

Abstract

Being one of the most complex orthopedic injuries, femur fractures are the ones asso ciated with prolonged recovery, high pain rates, and a strong requirement for precise alignment and optimal healing. Current methods for reduction rely heavily on man ual manipulation and real-time fluoroscopy, often resulting in suboptimal alignment, increased radiation exposure, and prolonged procedural time. Despite technological advances in surgical robotics, few systems have addressed these shortcomings in a way that combines precision, automation, and safety. In response to these limita tions, this thesis presents an autonomous robotic system for femur fracture reduction that integrates machine learning for actuator noise compensation. This research proposes a novel robot-assisted intelligent system for fracture reduc tion based on a Stewart platform architecture, integrating real-time sensor feedback and data-driven pose correction. A 3D model of the fractured femur is reconstructed using 3D Slicer, and MATLAB algorithms analyze the bone geometry to identify key anatomical features, such as peaks and troughs. Due to the inherent inaccuracy of physics-based solutions derived from incremental encoder data and non-linear me chanical deflection not accounted for by inverse kinematics, a machine learning model is introduced to estimate and correct pose errors based on empirical IMU feedback. A machine learning model is trained using this data to predict optimal Stewart platform leg adjustments based on the required alignment. Preliminary results suggest that machine learning-enhanced FK improves pose tracking accuracy over traditional methods, creating a more robust and adaptive control framework for robotic orthopedic interventions. This research contributes to the advancement of autonomous tele-surgeries, offering a data-driven approach to complex bone alignment procedures with the potential for minimal to zero surgeon assistance.